SECT. 2] CONTINENTAL SHELF AND SLOPE 305 



troughs at right angles to the direction of the coast. The problem is whether 

 these features can be explained only or essentially by glacial erosion. 



The glacial, or, more generally, erosional, explanation for the longitudinal 

 depressions meets with a serious difficulty in that the depressions are per- 

 pendicular to the normal flow of ice, or of rivers before and between the glacia- 

 tions. When the shelves into which they are incised bore ice-caps, it is evident 

 that the ice flowed through these depressions ; but it is not likely to have 

 deepened and enlarged them much because of their direction, whereas the over- 

 deepening was much easier in the transverse troughs, running in the general 

 trend of the ice flow. In other words, both the longitudinal and transverse 

 depressions may be initially of structural origin ; but the present forms prob- 

 ably owe much more to ice than to structure when they are transverse, and 

 much more to structure than to ice when they are longitudinal. 



The explanation put forward by O. and H. Holtedahl is that the longitudinal 

 depressions result from rifting which occurred during a late Tertiary uplift of 

 tiie continental margins. It agrees well with the evolution generally accepted 

 for Norway and several other arctic lands from which marginal depressions are 

 reported, but, as Kuenen and the writer separately pointed out, the coincidence 

 of the fissured shelves with glaciated areas is curious, and one wonders whether 

 there is any structural relation between the marginal fissures and the growing 

 and melting of ice-caps. The same idea is expressed by Jivago and Lissitzin, 

 who explain the facts observed in East Antarctica by a series of vertical iso- 

 static movements resulting from modifications in ice load. The main marginal 

 depression would be a hinge line allowing vertical "breathing" of the continent, 

 alternately depressed and uplifted relative to the deep-sea floor, the con- 

 tinental slope and even the outer shelf. A rifting on a smaller scale, related to 

 the same events, would account for smaller fissures. 



Would it not be possible to combine the two explanations? A Tertiary uplift 

 might have predisj^osed the shelves to be faulted ; further, this predisposition 

 might have led to marginal rift depressions in glaciated areas, owing to the iso- 

 static breathing of the continental margin, whereas in other regions flexiu-ing 

 or step-faulting generally occurred instead of rifting. If this interpretation were 

 right, it might be expected that the longitudinal rifts would not extend down 

 to the continental slope, since the ice-caps did not overload this part of the 

 continental margin. Here, however, a difficulty is found. The proflles across the 

 continental slope of the Norwegian Sea, which have been published and dis- 

 cussed by H. Holtedahl (1955) and have already been referred to, show distinct 

 rising blocks breaking the slope in steps. Thus, it appears as though the rifting 

 affecting the shelf has also affected the slope down to a depth of 700 fm or 

 more. In East Antarctica too, the description of the slope which has been 

 summarized above suggests evidence for rifting or block-faulting on various 

 scales in many areas lower than the level at which the Pleistocene Antarctic ice- 

 caps are likely to have rested upon the bottom. But this difficulty may be 

 solved if we consider that the peripheral belt of the area depressed under the 

 weight of the ice-cap may have been somewhat bulged upwards during the 



